Reaction path ensemble of the B-Z-DNA transition: a comprehensive atomistic study.

Since its discovery in 1979, left-handed Z-DNA has evolved from an in vitro curiosity to a challenging DNA structure with crucial roles in gene expression, regulation and recombination. A fundamental question that has puzzled researchers for decades is how the transition from B-DNA, the prevalent right-handed form of DNA, to Z-DNA is accomplished. Due to the complexity of the B-Z-DNA transition, experimental and computational studies have resulted in several different, apparently contradictory models. Here, we use molecular dynamics simulations coupled with state-of-the-art enhanced sampling techniques operating through non-conventional reaction coordinates, to investigate the B-Z-DNA transition at the atomic level. Our results show a complex free energy landscape, where several phenomena such as over-stretching, unpeeling, base pair extrusion and base pair flipping are observed resulting in interconversions between different DNA conformations such as B-DNA, Z-DNA and S-DNA. In particular, different minimum free energy paths allow for the coexistence of different mechanisms (such as zipper and stretch-collapse mechanisms) that previously had been proposed as independent, disconnected models. We find that the B-Z-DNA transition--in absence of other molecular partners--can encompass more than one mechanism of comparable free energy, and is therefore better described in terms of a reaction path ensemble.

Binding polymorphism in the DNA bound state of the Pdx1 homeodomain.

The subtle effects of DNA-protein recognition are illustrated in the homeodomain fold. This is one of several small DNA binding motifs that, in spite of limited DNA binding specificity, adopts crucial, specific roles when incorporated in a transcription factor. The homeodomain is composed of a 3-helix domain and a mobile N-terminal arm. Helix 3 (the recognition helix) interacts with the DNA bases through the major groove, while the N-terminal arm becomes ordered upon binding a specific sequence through the minor groove. Although many structural studies have characterized the DNA binding properties of homeodomains, the factors behind the binding specificity are still difficult to elucidate. A crystal structure of the Pdx1 homeodomain bound to DNA (PDB 2H1K) obtained previously in our lab shows two complexes with differences in the conformation of the N-terminal arm, major groove contacts, and backbone contacts, raising new questions about the DNA recognition process by homeodomains. Here, we carry out fully atomistic Molecular Dynamics simulations both in crystal and aqueous environments in order to elucidate the nature of the difference in binding contacts. The crystal simulations reproduce the X-ray experimental structures well. In the absence of crystal packing constraints, the differences between the two complexes increase during the solution simulations. Thus, the conformational differences are not an artifact of crystal packing. In solution, the homeodomain with a disordered N-terminal arm repositions to a partially specific orientation. Both the crystal and aqueous simulations support the existence of different stable binding conformers identified in the original crystallographic data with different degrees of specificity. We propose that protein-protein and protein-DNA interactions favor a subset of the possible conformations. This flexibility in DNA binding may facilitate multiple functions for the same transcription factor.

Theoretical modeling of spin crossover in metal-organic frameworks: [Fe(pz)2Pt(CN)4] as a case study.

Metal-organic frameworks (MOFs) with spin-crossover behavior are promising materials for applications in memory storage and sensing devices. A key parameter that characterizes these materials is the transition temperature T1/2, defined as the temperature with equal populations of low-spin and high-spin species. In this study, we describe the development, implementation, and application of a novel hybrid Monte Carlo/molecular dynamics method that builds upon the Ligand Field Molecular Mechanics approach and enables the modeling of spin-crossover properties in bulk materials. The new methodology is applied to the study of a spin-crossover MOF with molecular formula [Fe(pz)2Pt(CN)4] (pz = pyrazine). The total magnetic moment of the material is determined as a function of the temperature from direct calculations of the relative equilibrium populations of both low-spin and high-spin states of each Fe(II) center of the framework. The T1/2 value, calculated from the temperature dependence of the magnetization curve, is in good agreement with the available experimental data. A comparison between the spin-crossover behavior of the isolated secondary building block of the framework and the bulk material is presented, which reveals the origin of the different spin-crossover properties of the isolated molecular system and corresponding MOF structure.

Are long-range structural correlations behind the aggregration phenomena of polyglutamine diseases?

We have characterized the conformational ensembles of polyglutamine Qn peptides of various lengths n (ranging from 6 to 40), both with and without the presence of a C-terminal polyproline hexapeptide. For this, we used state-of-the-art molecular dynamics simulations combined with a novel statistical analysis to characterize the various properties of the backbone dihedral angles and secondary structural motifs of the glutamine residues. For Q40 (i.e., just above the pathological length ≃36 for Huntington's disease), the equilibrium conformations of the monomer consist primarily of disordered, compact structures with non-negligible α-helical and turn content. We also observed a relatively small population of extended structures suitable for forming aggregates including ß- and α-strands, and ß- and α-hairpins. Most importantly, for Q40 we find that there exists a long-range correlation (ranging for at least 20 residues) among the backbone dihedral angles of the Q residues. For polyglutamine peptides below the pathological length, the population of the extended strands and hairpins is considerably smaller, and the correlations are short-range (at most 5 residues apart). Adding a C-terminal hexaproline to Q40 suppresses both the population of these rare motifs and the long-range correlation of the dihedral angles. We argue that the long-range correlation of the polyglutamine homopeptide, along with the presence of these rare motifs, could be responsible for its aggregation phenomena.

The water hexamer: cage, prism, or both. Full dimensional quantum simulations say both.

State-of-the-art quantum simulations on a full-dimensional ab initio potential energy surface are used to characterize the properties of the water hexamer. The relative populations of the different isomers are determined over a wide range of temperatures. While the prism isomer is identified as the global minimum-energy structure, the quantum simulations, which explicitly include zero-point energy and quantum thermal motion, predict that both the cage and prism isomers are present at low temperature down to almost 0 K. This is largely consistent with the available experimental data and, in particular, with very recent measurements of broadband rotational spectra of the water hexamer recorded in supersonic expansions.

Conformations and free energy landscapes of polyproline peptides.

The structure of the proline amino acid allows folded polyproline peptides to exist as both left- (PPII) and right-handed (PPI) helices. We have characterized the free energy landscapes of hexamer, nanomer, and tridecamer polyproline peptides in gas phase and implicit water as well as explicit hexane and 1-propanol for the nanomer. To enhance the sampling provided by regular molecular dynamics, we used the recently developed adaptively biased molecular dynamics method, which describes Landau free energy maps in terms of relevant collective variables. These maps, as a function of the collective variables of handedness, radius of gyration, and three others based on the peptide torsion angle omega, were used to determine the relative stability of the different structures, along with an estimate of the transition pathways connecting the different minima. Results show the existence of several metastable isomers and therefore provide a complementary view to experimental conclusions based on photo-induced electron transfer experiments with regard to the existence of stable heterogeneous subpopulations in PPII polyproline.

A statistical analysis of the PPII propensity of amino acid guests in proline-rich peptides.

There has been considerable debate about the intrinsic PPII propensity of amino-acid residues in denatured polypeptides. Experimentally, the propensity scale is based on the behavior of guest amino-acid residues placed in the middle of polyproline hosts. We have used classical molecular dynamics simulations, with state-of-the-art force fields to carry out a comprehensive analysis of the conformational equilibria of the proline-based host oligopeptides with single guests. The tracked structural characteristics include the PPII content, the cis/trans isomerization of the prolyl bonds, the puckering of the pyrrolidine rings of the proline residues, and the secondary structural motifs. We find no evidence for an intrinsic PPII propensity in any of the guest amino acids other than proline. Instead, the PPII content as derived from experiments may be explained in terms of: 1), a local correlation between the dihedral angles of the guest amino acid and the proline residue immediately preceding it; and 2), a nonlocal correlation between the cis/trans states of the peptide bonds. In terms of the latter, we find that the presence of a guest (other than proline, tyrosine, or tryptophan) increases the trans content of most of the prolyl bonds, which results in an effective increase of the peptide PPII content. With respect to the local dihedral correlations, we find that these are well described in terms of the so-called odds-ratio statistic. Expressed in terms of free energy language, the PPII content based on the odds-ratio of the relevant residues correlate well with the experimentally measured PPII content.

The α-sheet: a missing-in-action secondary structure?

The α-sheet has been speculated to play a role as a toxic conformer in amyloid diseases. However, except for relatively short fragments, its detection has remained elusive. Here, we present molecular dynamics simulations that support the existence of the α-sheet as a stable, metastable, or long-lived secondary structure in polyglutamine and, to a lesser extent, in polyasparagine aggregates.

Conformational free energies of methyl-alpha-L-iduronic and methyl-beta-D-glucuronic acids in water.

We present a simulation protocol that allows for efficient sampling of the degrees of freedom of a solute in explicit solvent. The protocol involves using a nonequilibrium umbrella sampling method, in this case, the recently developed adaptively biased molecular dynamics method, to compute an approximate free energy for the slow modes of the solute in explicit solvent. This approximate free energy is then used to set up a Hamiltonian replica exchange scheme that samples both from biased and unbiased distributions. The final accurate free energy is recovered via the weighted histogram analysis technique applied to all the replicas, and equilibrium properties of the solute are computed from the unbiased trajectory. We illustrate the approach by applying it to the study of the puckering landscapes of the methyl glycosides of alpha-L-iduronic acid and its C5 epimer beta-D-glucuronic acid in water. Big savings in computational resources are gained in comparison to the standard parallel tempering method.

A classical molecular dynamics investigation of the free energy and structure of short polyproline conformers.

Folded polyproline peptides can exist as either left-(PPII) or right-handed (PPI) helices, depending on their environment. In this work, we have characterized the conformations and the free energy landscapes of Ace-(Pro)(n)-Nme, n=2,3, ... ,9, and 13 peptides both in vacuo and in an implicit solvent environment. In order to enhance the sampling provided by regular molecular dynamics simulations, we have used the recently developed adaptively biased molecular dynamics method--which provides an accurate description of the free energy landscapes in terms of a set of relevant collective variables--combined with Hamiltonian and temperature replica exchange molecular dynamics methods. The collective variables, which are chosen so as to reflect the stable structures and the "slow modes" of the polyproline system, were based primarily on properties of length and of the cis/trans isomerization associated with the prolyl bonds. Results indicate that the space of peptide structures is characterized not just by pure PPII and PPI structures, but rather by a broad distribution of stable minima with similar free energies. These results are in agreement with recent experimental work. In addition, we have used steered molecular dynamics methods in order to quantitatively estimate the free energy difference of PPI and PPII for peptides of the length n=2, ... ,5 in vacuo and implicit water and qualitatively investigate transition pathways and mechanisms for the PPII to PPI transitions. A zipper-like mechanism, starting from either the center of the peptide or the amidated end, appear to be the most likely mechanisms for the PPII→PPI transition for the longer peptides.

A New Mixed All-Atom/Coarse-Grained Model: Application to Melittin Aggregation in Aqueous Solution.

We introduce a new mixed resolution, all-atom/coarse-grained approach (AACG), for modeling peptides in aqueous solution and apply it to characterizing the aggregation of melittin. All of the atoms in peptidic components are represented, while a single site is used for each water molecule. With the full flexibility of the peptide retained, our AACG method achieves speedups by a factor of 3-4 for CPU time reduction and another factor of roughly 7 for diffusion. An Ewald treatment permits the inclusion of long-range electrostatic interactions. These characteristics fit well with the requirements for studying peptide association and aggregation, where the system sizes and time scales require considerable computational resources with all-atom models. In particular, AACG is well suited for biologics since changes in peptide shape and long-range electrostatics may play an important role. The application of AACG to melittin, a 26-residue peptide with a well-known propensity to aggregate in solution, serves as an initial demonstration of this technology for studying peptide aggregation. We observed the formation of melittin aggregates during our simulations and characterized the time-evolution of aggregate size distribution, buried surface areas, and residue contacts. Key interactions including π-cation and π-stacking involving TRP19 were also examined. Our AACG simulations demonstrated a clear salt effect and a moderate temperature effect on aggregation and support the molten globule model of melittin aggregates. As a showcase, this work illustrates the useful role for AACG in investigations of peptide aggregation and its potential to guide formulation and design of biologics.

Molecular dynamics simulations of DNA with polarizable force fields: convergence of an ideal B-DNA structure to the crystallographic structure.

We have investigated to what extent molecular dynamics (MD) simulations can reproduce DNA sequence-specific features, given different electrostatic descriptions and different cell environments. For this purpose, we have carried out multiple unrestrained MD simulations of the DNA duplex d(CCAACGTTGG)2. With respect to the electrostatic descriptions, two different force fields are studied: a traditional description based on atomic point charges and a polarizable force field. With respect to the cell environment, the difference between crystal and solution environments is emphasized, as well as the structural importance of divalent ions. By imposing the correct experimental unit cell environment, an initial configuration with two ideal B-DNA duplexes in the unit cell is shown to converge to the crystallographic structure. This convergence is measured by the appearance of sequence-dependent features that very closely resemble the crystallographic ones as well as by the decay of the all-atom root-mean-squared coordinates deviations (RMSD) with respect to the crystallographic structure. Given the appropriate crystallographic constraints, this is the first example of multiple nanosecond molecular dynamics trajectory that shows an ideal B-DNA model converging to an experimental structure, with a significant decay of RMSD.

Deprotonation of solvated formic acid: Car-Parrinello and metadynamics simulations.

The deprotonation of solvated formic acid was investigated theoretically with ab initio simulations. With the Car-Parrinello method, deprotonation and reprotonation by means of a proton wire were observed. The microscopics of these reactions were analyzed, and reveal the key role played by nearby water molecules in catalyzing the reactions. A constrained molecular dynamics calculation was carried out to estimate the dissociation free energy. Deprotonation of formic acid was further investigated with the recently developed metadynamics method using the formic acid oxygen coordination numbers as the collective variables. The determined free-energy landscape gives barriers similar to that obtained with the constrained free-energy calculation.

Minimization of the Renyi entropy production in the stationary states of the Brownian process with matched death and birth rates.

We analyze the Fleming-Viot process. The system is confined in a box, whose boundaries act as a sink of Brownian particles. The death rate at the boundaries is matched by the branching (birth) rate in the system and thus the number of particles is kept constant. We show that such a process is described by the Renyi entropy whose production is minimized in the stationary state. The entropy production in this process is a monotonically decreasing function of time irrespective of the initial conditions. The first Laplacian eigenvalue is shown to be equal to the Renyi entropy production in the stationary state. As an example we simulate the process in a two-dimensional box.

Development of a "First Principles" Water Potential with Flexible Monomers. II: Trimer Potential Energy Surface, Third Virial Coefficient, and Small Clusters.

A full-dimensional potential energy function (MB-pol) for simulations of water from the dimer to bulk phases is developed entirely from "first principles" by building upon the many-body expansion of the interaction energy. Specifically, the MB-pol potential is constructed by combining a highly accurate dimer potential energy surface [J. Chem. Theory Comput. 2013, 9, 5395] with explicit three-body and many-body polarization terms. The three-body contribution, expressed as a combination of permutationally invariant polynomials and classical polarizability, is derived from a fit to more than 12000 three-body energies calculated at the CCSD(T)/aug-cc-pVTZ level of theory, imposing the correct asymptotic behavior as predicted from "first principles". Here, the accuracy of MB-pol is demonstrated through comparison of the calculated third virial coefficient with the corresponding experimental data as well as through analysis of the relative energy differences of small clusters.

Development of a "First-Principles" Water Potential with Flexible Monomers. III. Liquid Phase Properties.

The MB-pol full-dimensional water potential introduced in the first two papers of this series [J. Chem. Theory Comput. 2013, 9, 5395 and J. Chem. Theory Comput. 2014, 10, 1599] is employed here in classical and quantum simulations of liquid water under ambient conditions. Comparisons with the available experimental data for several structural, thermodynamic, and dynamical properties indicate that MB-pol provides a highly accurate description of the liquid phase. Combined with previous analyses of the dimer vibration-rotation tunneling spectrum, second and third virial coefficients, and cluster structures and energies, the present results demonstrate that MB-pol represents a major step toward the long-sought "universal model" capable of describing the properties of water from the gas to the condensed phases.

A Critical Assessment of Two-Body and Three-Body Interactions in Water.

The microscopic behavior of water under different conditions and in different environments remains the subject of intense debate. A great number of the controversies arise due to the contradictory predictions obtained within different theoretical models. Relative to conclusions derived from force fields or density functional theory, there is comparably less room to dispute highly correlated electronic structure calculations. Unfortunately, such ab initio calculations are severely limited by system size. In this study, a detailed analysis of the two- and three-body water interactions evaluated at the CCSD(T) level is carried out to quantitatively assess the accuracy of several force fields, DFT models, and ab initio based interaction potentials that are commonly used in molecular simulations. On the basis of this analysis, a new model, HBB2-pol, is introduced which is capable of accurately mapping CCSD(T) results for water dimers and trimers into an efficient analytical function. The accuracy of HBB2-pol is further established through comparison with the experimentally determined second and third virial coefficients.

Development of a "First Principles" Water Potential with Flexible Monomers: Dimer Potential Energy Surface, VRT Spectrum, and Second Virial Coefficient.

The development of a "first principles" water potential with flexible monomers (MB-pol) for molecular simulations of water systems from gas to condensed phases is described. MB-pol is built upon the many-body expansion of the intermolecular interactions, and the specific focus of this study is on the two-body term (V2B) representing the full-dimensional intermolecular part of the water dimer potential energy surface. V2B is constructed by fitting 40,000 dimer energies calculated at the CCSD(T)/CBS level of theory and imposing the correct asymptotic behavior at long-range as predicted from "first principles". The comparison of the calculated vibration-rotation tunneling (VRT) spectrum and second virial coefficient with the corresponding experimental results demonstrates the accuracy of the MB-pol dimer potential energy surface.

Electrostatics interactions in classical simulations.

Electrostatic interactions are crucial for both the accuracy and performance of atomistic biomolecular simulations. In this chapter we review well-established methods and current developments aiming at efficiency and accuracy. Specifically, we review the classical Ewald summations, particle-particle particle-method particle-method Ewald algorithms, multigrid, fast multipole, and local methods. We also highlight some recent developments targeting more accurate, yet classical, representation of the molecular charge distribution.

Recipes for free energy calculations in biomolecular systems.

During the last decade, several methods for sampling phase space and calculating various free energies in biomolecular systems have been devised or refined for molecular dynamics (MD) simulations. Thus, state-of-the-art methodology and the ever increasing computer power allow calculations that were forbidden a decade ago. These calculations, however, are not trivial as they require knowledge of the methods, insight into the system under study, and, quite often, an artful combination of different methodologies in order to avoid the various traps inherent in an unknown free energy landscape. In this chapter, we illustrate some of these concepts with two relatively simple systems, a sugar ring and proline oligopeptides, whose free energy landscapes still offer considerable challenges. In order to explore the configurational space of these systems, and to surmount the various free energy barriers, we combine three complementary methods: a nonequilibrium umbrella sampling method (adaptively biased MD, or ABMD), replica-exchange molecular dynamics (REMD), and steered molecular dynamics (SMD). In particular, ABMD is used to compute the free energy surface of a set of collective variables; REMD is used to improve the performance of ABMD, to carry out sampling in space complementary to the collective variables, and to sample equilibrium configurations directly; and SMD is used to study different transition mechanisms.